1/*-
2 * SPDX-License-Identifier: BSD-2-Clause
3 *
4 * Copyright (c) 2019 Yandex LLC
5 * Copyright (c) 2019 Andrey V. Elsukov <ae@FreeBSD.org>
6 * Copyright (c) 2019 Boris N. Lytochkin <lytboris@gmail.com>
7 *
8 * Redistribution and use in source and binary forms, with or without
9 * modification, are permitted provided that the following conditions
10 * are met:
11 *
12 * 1. Redistributions of source code must retain the above copyright
13 *    notice, this list of conditions and the following disclaimer.
14 * 2. Redistributions in binary form must reproduce the above copyright
15 *    notice, this list of conditions and the following disclaimer in the
16 *    documentation and/or other materials provided with the distribution.
17 *
18 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR
19 * IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
20 * OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
21 * IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT,
22 * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
23 * NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
24 * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
25 * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
26 * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF
27 * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
28 */
29
30#include <sys/param.h>
31#include <sys/systm.h>
32#include <sys/counter.h>
33#include <sys/kernel.h>
34#include <sys/lock.h>
35#include <sys/mbuf.h>
36#include <sys/module.h>
37#include <sys/rmlock.h>
38#include <sys/rwlock.h>
39#include <sys/socket.h>
40#include <sys/sysctl.h>
41
42#include <net/if.h>
43#include <net/if_var.h>
44#include <net/if_pflog.h>
45#include <net/pfil.h>
46
47#include <netinet/in.h>
48#include <netinet/ip.h>
49#include <netinet/ip_icmp.h>
50#include <netinet/ip_var.h>
51#include <netinet/ip_fw.h>
52#include <netinet/ip6.h>
53#include <netinet/icmp6.h>
54#include <netinet6/ip_fw_nat64.h>
55
56#include <netpfil/ipfw/ip_fw_private.h>
57#include <netpfil/pf/pf.h>
58
59#include "nat64clat.h"
60
61#define	NAT64_LOOKUP(chain, cmd)	\
62	(struct nat64clat_cfg *)SRV_OBJECT((chain), (cmd)->arg1)
63
64static void
65nat64clat_log(struct pfloghdr *plog, struct mbuf *m, sa_family_t family,
66    uint32_t kidx)
67{
68	static uint32_t pktid = 0;
69
70	memset(plog, 0, sizeof(*plog));
71	plog->length = PFLOG_HDRLEN;
72	plog->af = family;
73	plog->action = PF_NAT;
74	plog->dir = PF_IN;
75	plog->rulenr = htonl(kidx);
76	pktid++;
77	plog->subrulenr = htonl(pktid);
78	plog->ruleset[0] = '\0';
79	strlcpy(plog->ifname, "NAT64CLAT", sizeof(plog->ifname));
80	ipfw_bpf_mtap2(plog, PFLOG_HDRLEN, m);
81}
82
83static int
84nat64clat_handle_ip4(struct ip_fw_chain *chain, struct nat64clat_cfg *cfg,
85    struct mbuf *m)
86{
87	struct pfloghdr loghdr, *logdata;
88	struct in6_addr saddr, daddr;
89	struct ip *ip;
90
91	ip = mtod(m, struct ip*);
92	/* source address for CLAT may be private with no harm */
93	if (nat64_check_ip4(ip->ip_src.s_addr) != 0 ||
94	    nat64_check_ip4(ip->ip_dst.s_addr) != 0 ||
95	    nat64_check_private_ip4(&cfg->base, ip->ip_dst.s_addr) != 0)
96		return (NAT64SKIP);
97
98	memcpy(&saddr, &cfg->base.clat_prefix, sizeof(saddr));
99	nat64_embed_ip4(&saddr, cfg->base.clat_plen, ip->ip_src.s_addr);
100	memcpy(&daddr, &cfg->base.plat_prefix, sizeof(daddr));
101	nat64_embed_ip4(&daddr, cfg->base.plat_plen, ip->ip_dst.s_addr);
102	if (cfg->base.flags & NAT64_LOG) {
103		logdata = &loghdr;
104		nat64clat_log(logdata, m, AF_INET, cfg->no.kidx);
105	} else
106		logdata = NULL;
107	return (nat64_do_handle_ip4(m, &saddr, &daddr, 0, &cfg->base,
108	    logdata));
109}
110
111static int
112nat64clat_handle_ip6(struct ip_fw_chain *chain, struct nat64clat_cfg *cfg,
113    struct mbuf *m)
114{
115	struct pfloghdr loghdr, *logdata;
116	struct ip6_hdr *ip6;
117	uint32_t aaddr;
118
119	/*
120	 * NOTE: we expect ipfw_chk() did m_pullup() up to upper level
121	 * protocol's headers. Also we skip some checks, that ip6_input(),
122	 * ip6_forward(), ip6_fastfwd() and ipfw_chk() already did.
123	 */
124	ip6 = mtod(m, struct ip6_hdr *);
125	/* Check ip6_dst matches configured prefix */
126	if (memcmp(&ip6->ip6_dst, &cfg->base.clat_prefix,
127	    cfg->base.clat_plen / 8) != 0)
128		return (NAT64SKIP);
129	/* Check ip6_src matches configured prefix */
130	if (memcmp(&ip6->ip6_src, &cfg->base.plat_prefix,
131	    cfg->base.plat_plen / 8) != 0)
132		return (NAT64SKIP);
133
134	if (cfg->base.flags & NAT64_LOG) {
135		logdata = &loghdr;
136		nat64clat_log(logdata, m, AF_INET6, cfg->no.kidx);
137	} else
138		logdata = NULL;
139
140	aaddr = nat64_extract_ip4(&ip6->ip6_src, cfg->base.plat_plen);
141	return (nat64_do_handle_ip6(m, aaddr, 0, &cfg->base, logdata));
142}
143
144static int
145nat64clat_handle_icmp6(struct ip_fw_chain *chain, struct nat64clat_cfg *cfg,
146    struct mbuf *m)
147{
148	struct pfloghdr loghdr, *logdata;
149	struct nat64_counters *stats;
150	struct ip6_hdr *ip6i;
151	struct icmp6_hdr *icmp6;
152	uint32_t daddr;
153	int hlen, proto;
154
155	hlen = 0;
156	stats = &cfg->base.stats;
157	proto = nat64_getlasthdr(m, &hlen);
158	if (proto != IPPROTO_ICMPV6) {
159		NAT64STAT_INC(stats, dropped);
160		return (NAT64MFREE);
161	}
162	icmp6 = mtodo(m, hlen);
163	switch (icmp6->icmp6_type) {
164	case ICMP6_DST_UNREACH:
165	case ICMP6_PACKET_TOO_BIG:
166	case ICMP6_TIME_EXCEED_TRANSIT:
167	case ICMP6_PARAM_PROB:
168		break;
169	default:
170		NAT64STAT_INC(stats, dropped);
171		return (NAT64MFREE);
172	}
173	hlen += sizeof(struct icmp6_hdr);
174	if (m->m_pkthdr.len < hlen + sizeof(struct ip6_hdr) + ICMP_MINLEN) {
175		NAT64STAT_INC(stats, dropped);
176		return (NAT64MFREE);
177	}
178	if (m->m_len < hlen + sizeof(struct ip6_hdr) + ICMP_MINLEN)
179		m = m_pullup(m, hlen + sizeof(struct ip6_hdr) + ICMP_MINLEN);
180	if (m == NULL) {
181		NAT64STAT_INC(stats, nomem);
182		return (NAT64RETURN);
183	}
184	/*
185	 * Use destination address from inner IPv6 header to determine
186	 * IPv4 mapped address.
187	 */
188	ip6i = mtodo(m, hlen);
189	daddr = nat64_extract_ip4(&ip6i->ip6_dst, cfg->base.clat_plen);
190	if (daddr == 0) {
191		NAT64STAT_INC(stats, dropped);
192		return (NAT64MFREE);
193	}
194	if (cfg->base.flags & NAT64_LOG) {
195		logdata = &loghdr;
196		nat64clat_log(logdata, m, AF_INET6, cfg->no.kidx);
197	} else
198		logdata = NULL;
199	return (nat64_handle_icmp6(m, 0, daddr, 0, &cfg->base, logdata));
200}
201
202int
203ipfw_nat64clat(struct ip_fw_chain *chain, struct ip_fw_args *args,
204    ipfw_insn *cmd, int *done)
205{
206	ipfw_insn *icmd;
207	struct nat64clat_cfg *cfg;
208	int ret;
209
210	IPFW_RLOCK_ASSERT(chain);
211
212	*done = 0; /* try next rule if not matched */
213	icmd = cmd + 1;
214	if (cmd->opcode != O_EXTERNAL_ACTION ||
215	    cmd->arg1 != V_nat64clat_eid ||
216	    icmd->opcode != O_EXTERNAL_INSTANCE ||
217	    (cfg = NAT64_LOOKUP(chain, icmd)) == NULL)
218		return (0);
219
220	switch (args->f_id.addr_type) {
221	case 4:
222		ret = nat64clat_handle_ip4(chain, cfg, args->m);
223		break;
224	case 6:
225		ret = nat64clat_handle_ip6(chain, cfg, args->m);
226		break;
227	default:
228		return (0);
229	}
230
231	if (ret == NAT64SKIP) {
232		/*
233		 * In case when packet is ICMPv6 message from an intermediate
234		 * router, the source address of message will not match the
235		 * addresses from configured prefixes.
236		 */
237		if (args->f_id.proto != IPPROTO_ICMPV6)
238			return (0);
239
240		ret = nat64clat_handle_icmp6(chain, cfg, args->m);
241	}
242
243	if (ret == NAT64SKIP)
244		return (0);
245
246	*done = 1; /* terminate the search */
247	if (ret == NAT64MFREE)
248		m_freem(args->m);
249
250	args->m = NULL;
251	return (IP_FW_NAT64);
252}
253